Reproductive Physiology of the Female Cat

(1)

Reproductive Physiology of the Female Cat

With special reference to cervical patency, sperm distribution and hysterography

Kaywalee Chatdarong

Department of Obstetrics and Gynaecology Uppsala

Doctoral thesis

Swedish University of Agricultural Sciences

Uppsala 2003

(2)

Acta Universitatis Agriculturae Sueciae Veterinaria 162

ISSN 1401-6257 ISBN 91-576-6395-5

(3)

Abstract

Chatdarong, K. 2003. Reproductive physiology of the female cat, with special reference to cervical patency, sperm distribution and hysterography. Doctoral dissertation.

ISSN: 1401-6257, ISBN: 91-576-6395-5.

Ovarian cyclicity induces considerable changes in morphology, physiology and function of the reproductive organs in animals. This thesis aimed to study reproductive physiology of the female domestic cat, focusing on the cervix, uterus and uterine tube during different stages of the oestrous cycle, after mating and under pathological conditions. Transcervical catheterisation was performed using a specially designed catheter. Patency of the cervix and uterine motility were studied with the aids of fluoroscopy and scintigraphy by depositing contrast fluids and radiopharmaceutical medium in the cranial vagina. The relationship between cervical patency and oestrous behaviour, cornification of the vaginal cells and serum concentrations of oestradiol-17β were evaluated. Hysterography, a positive contrast study, was performed to illustrate the uterine appearance during the oestrous cycle in relation to histological characteristics of the endometrium. Immunohistochemistry using proliferating nuclear antigen (PCNA) was introduced to identify mitotic activity of the endometrial cells. Patency of the cervix to spermatozoa as well as sperm distribution after natural mating was evaluated by flushing the vagina, the uterus and the uterine tube. For the study of sperm distribution in the female reproductive tract, two methods, flushing and tissue sectioning were evaluated and compared.

Using a specially designed catheter it was possible to catheterise the cervix during interoestrus, oestrus, metoestrus and postpartum and to introduce contrast fluids into the uterus to study uterine appearance using hysterography also when the cervix was closed.

The period when the cervix was patent was found to vary among individuals: the cervix was open either only during late-oestrus; during midoestrus and late-oestrus; or throughout the entire behavioural oestrus. Patency of the cervix was found to usually coincide with the maximum degree of vaginal cornification and thus, presumably is regulated at least in part by the serum concentration of oestradiol-17β. Hysterograms revealed differences in uterine luminal shape that corresponded to the histological characteristics of the endometrium in cats at various stages of the oestrous cycle, cats given medroxyprogesterone acetate (MPA) and cats with uterine pathology. Straight- and wavy-shaped uterine lumens were characteristic of the uterine horns in the inactive and follicular stages of the oestrous cycle.

Coil-shaped uterine lumen appeared to be a progestagenic effect seen in the luteal, the MPA-treated and the pathological groups. A coiled uterine cavity was suggestive of endometrial hyperplasia, whereas irregular filling defects were indicative of generalised cystic changes in the endometrium. The expression of PCNA in luminal and glandular epithelial cells was observed although the mitotic activity was not related to neither stages of oestrous cycle nor uterine pathological conditions. The results from the sperm distribution study demonstrated that the cervix and the uterotubal junction (UTJ) were sperm barriers in the cat. The endometrial crypts and the UTJ functioned as sperm reservoirs before ovulation, whereas the isthmus was a sperm reservoir around the time of ovulation.

The observations determined the dynamics of the cervix, the uterus and the uterine tube in the cat in relation to ovarian activity. This thesis is the first to provide nomenclatures for describing uterine appearance in hysterograms according to the shape of the uterine horns, luminal cavity and the characteristics of the intraluminal lining, to demonstrate the PCNA expression as well as to determine the distribution of spermatozoa in the female reproductive tract of the cat after natural mating.

Key words: feline, queen, copulation, oviduct, radiography, endometrial hyperplasia Author’s address: Kaywalee Chatdarong, Department of Obstetrics and Gynaecology, Faculty of Veterinary Medicine, Swedish University of Agricultural Sciences (SLU), PO Box 7039, SE-750 07 Uppsala, Sweden (On leave from the Department of Obstetrics, Gynaecology and Reproduction, Faculty of Veterinary Science, Chulalongkorn University, Bangkok 10330, Thailand). E.mail: Kaywalee.Chatdarong@og.slu.se

(4)

To my parents and my sister

(5)

Appendix

Papers I-IV

The present thesis is based on the following papers, which will be referred to by their Roman numerals:

I. Chatdarong, K., Lohachit, C., Ponglowhapan, S. & Linde-Forsberg, C. 2001.

Transcervical catheterization and cervical patency in domestic cats. Journal of Reproduction and Fertility Supplement 57, 353-356.

II. Chatdarong, K., Kampa, N., Axnér, E. & Linde-Forsberg, C. 2002.

Investigation of cervical patency and uterine appearance in domestic cats by fluoroscopy and scintigraphy. Reproduction in Domestic Animals 37, 275- 281.

III. Chatdarong, K., Rungsipipat, A., Axnér, E. & Linde-Forsberg, C.

Hysterographic appearance and uterine histology at different stages of the reproductive cycle and after progestagen treatment in the domestic cat.

Manuscript submitted for publication.

IV. Chatdarong, K., Lohachit, C. & Linde-Forsberg, C. Distribution of spermatozoa in the female reproductive tract of the domestic cat in relation to ovulation induced by natural mating. Manuscript submitted for publication.

Papers I and II are reproduced with permission of the journals concerned.

(8)

Abbreviations

ABC avidin-biotin peroxidase complex AI artificial insemination

CH corpora haemorrhagica CL corpora lutea

eCG equine chorionic gonadotrophin FSH follicle stimulating hormone FSH-P pituitary follicle stimulating hormone GnRH gonadotrophin-releasing hormone hCG human chorionic gonadotrophin HSA human serum albumin particle LH luteinising hormone

MPA medroxyprogesterone acetate PBS phosphate-buffered saline solution PCNA proliferative cell nuclear antigen PMN polymorphonuclear cell

99mTc technetium 99m UTJ uterotubal junction

(9)

Introduction

Background

The domestic cat is one of the most popular pets. Considerable emphasis on the study of reproduction in this species has been generated with the increasing interest in the domestic cat as a model for research, aimed at the preservation of endangered wild felids (Wildt et al., 1986), and as an experimental model for at least 36 human physiological abnormalities, including inherited genetic disorders (Goodrowe et al., 1989). The cat is also important for studying the phenomenon of copulation-induced ovulation, a conserved mechanism. Except for the domestic cat (Felis catus), the Northern European lynx (Lynx lynx) and a few other felids, most of the 36 felid species are classified as threatened, vulnerable or endangered by the Convention on International Trade in Endangered Species (CITES, 2000).

In Sweden, the only wild felid - the lynx- is classed as vulnerable by the Swedish Red List of threatened species. The population is estimated to be between 1400 and 1800 individuals and is controlled by strictly regulated hunting to keep the number of animals to a level where they can be tolerated by the public. This number of individuals is close to the minimum that is estimated to be necessary for a favourable development of this species in Sweden (Swedish Environmental Protection Agency, Dnr 411-6644-01). Among nine species of wild felids in Thailand, one (Felis marmorata) is considered nearly extinct. The other eight species (Panthera tigris, Panthera pardus, Neofelis nebulosa, Felis temmincki, Felis viverrina, Felis bengalensis, Felis chaus and Felis planiceps)are considered at risk of extinction.

Despite the growing popularity of the cat as a pet, most of the breeding of pedigree cats takes place within small catteries (Farstad, 2000) and a high percentage of male cats are castrated at an early age. Thus, inbreeding is a major problem both in the small populations of wild felid species and in purebred domestic cats (Axnér, 2000). Improved knowledge of the basic reproductive physiology of the female cat is instrumental for taking advantage of the techniques for assisted reproduction, and invaluable in clinical practice and diagnosis of uterine pathologies in this species.

Oestrous cycle

Ovarian activity in the female cat is dependent on photoperiod. The cat is polyoestrus and also an induced ovulator (Shille et al., 1979). Oestrus in free- ranging females is induced by an increase in day length, whereas decreasing photoperiod results in seasonal anoestrus. However, high ambient temperatures during summer may reduce the incidence of oestrus (Concannon & Lein, 1983;

Feldman & Nelson, 1996). Ovarian inactivity during winter anoestrus is related to an elevation in circulating prolactin and melatonin in peripheral plasma (Leyva et al., 1984; Leyva et al., 1989). In cats maintained indoors, ovarian activity can be controlled by artificial light. A minimum of 10-hr artificial light, equivalent to that

(10)

of 100-watt bulb in a 4 x 4 m room, can provide oestrous cycles throughout the year (Shille & Sojka, 1995). It is possible to induce oestrus by social stimuli by either a tomcat or an oestrous female (Michel, 1993).

The female cat usually enters the follicular phase abruptly. Oestrus may be preceded by a 1-2 day period of pro-oestrus characterised by female oestrous behavioural signs but not male acceptance. The follicles develop to the vesicular stage within 24 to 48 hr prior to the first day of oestrous behaviour (Wildt &

Seager, 1980). The oestrous stage ranges from 3 to 16 days (average 7 days).

Behavioural changes occur often overnight from interoestrus or anoestrus and may precede changes in vaginal cornification (Shille & Sojka, 1995). A rapid growth of the follicles results in a two-fold increase in the peripheral plasma concentration of oestradiol-17β within 24 hr (Shille et al., 1979). If the queen is not mated, ovulation will not occur in most cases and the preovulatory follicles become atretic and degenerated. The cats undergo an interoestrous cyclic stage, which lasts from 2 to 19 days (average 7 days), characterised by the absence of oestrous behaviour and with basal levels of plasma oestradiol-17β concentrations (Shille &

Sojka, 1995). In some oestrous periods, overlapping follicular waves cause a continuous follicular phase (Feldman & Nelson, 1996).

Artificial induction of oestrus

Pituitary follicular stimulating hormone (FSH-P) and equine chorionic gonadotrophin (eCG) are commonly used to induce follicular development in the cat. One administration of 100 iu eCG intramuscularly to anoestrous cats, followed 5-7 days later by an injection of 50 iu human chorionic gonadotrophin (hCG), produces ovulation and pregnancy results comparable to those of natural matings (Cline et al., 1980). The ovaries of the domestic cat are sensitive to over- dosing with gonadotrophins. High single doses (1000 iu) and multiple administrations of low-dose eCG result in premature luteinisation of early developing follicles (Wildt et al., 1978; Cline et al., 1980). Repeated injections of eCG may lead to the production of antibodies against FSH and luteinising hormone (LH), and a subsequent decreased response to stimulation or infertility (Swanson et al., 1995). FSH-P given in a 2-mg intramuscular daily dose for 5-7 days generally produces an oestrous behavioural response and follicular development (Wildt et al., 1978). Pregnancies have been obtained from females artificially inseminated with frozen-thawed semen following treatment with FSH- P (2 mg/day for 5 days) (Platz et al., 1978).

Mating and ovulation

A coital stimulus induces a neural firing reflex that stimulates the medial basal hypothalamus to synthesize and liberate gonadotrophin-releasing hormone (GnRH), which then stimulates the release of pituitary luteinising hormone (LH) (Robinson & Sawyer, 1987). Successful ovulation depends on the presence of mature follicles and an adequate surge of LH. Females require several days of oestrogen priming before copulation can induce a surge of LH sufficient to induce

(11)

ovulation. Although females are often sexually receptive by the second or third day of ovarian follicular growth, in some cats the amount of LH released in response to copulation may be limited until the animal reaches the fourth or fifth day of follicular growth (third or fourth day of oestrus) (Banks & Stabenfeldt, 1982). An adequate LH surge also depends on the number and frequency of copulations. LH release occurs rapidly with increases in serum concentrations evident within 5 min after mating (Johnson & Gay, 1981) and remaining elevated until 8 hr in ovulating cats, whereas the LH levels remain low in the cats that do not ovulate (Concannon et al., 1980).

Multiple copulation regimens such as four copulations during a 21-81 min period or ad libitum copulatory activity for 4-hrs (8-12 copulations) results in 100% successful ovulation whereas a single copulation induces an LH release sufficient to cause ovulation in only half of cats (Concannon et al., 1980). The greatest and the most prolonged LH releases are induced by ad libitum copulatory activity for 4-hrs. The intervals between copulations vary among individuals, being brief (6-12 min) shortly after the start of unrestricted male-female interaction and longer (18-61 min) toward the end of a 4-hr period (Concannon et al., 1980).

Although cats often copulate four times within the first hour, a single copulation on the third day of oestrus, when oestrous signs are most obvious and the oestradiol concentration is markedly high has been reported to be sufficient for an adequate LH release to induce ovulation (Tsutsui & Stabenfeldt, 1993). Ovulation occurs 25-32 hr after copulation (Shille et al., 1983) and can occur as late as 52 hr after the first LH peak (Wildt et al., 1981). Spontaneous ovulation can be observed in up to 35% of older cats housed individually with only visualization of other members in the colony (Lawler et al., 1993) and in young, group-housed cats (Gudermuth et al., 1997). The cats ovulate secondary oocytes in an all-or-none fashion depending on sufficient release of LH (Banks & Stabenfeldt, 1982).

After ovulation, the ovaries display reddish prominent corpora haemorrhagica (CH) characterised by high vascularisation. Thereafter, the developing copora lutea (CL) turn to pink or orange in colour and grow to 4.5 mm in diameter by days 12-16 after copulation (Dawson, 1946). Ovarian activity shifts from oestrogen to progesterone secretion 44-60 hr after the LH peak (Wildt et al., 1981). The progesterone concentration in peripheral plasma does not increase until 3-4 days after copulation (Verhage et al., 1976). CL of pregnancy remains functional throughout gestation, regressing at or near parturition (Verhage et al., 1976; Schmidt et al., 1983). However, in the nursing cat, the CLs remain well developed histologically until 63 days postpartum (Dawson, 1946).

Artificial induction of ovulation

Other than natural mating, stimuli, including direct vaginal stimulation with a vaginal swab or a glass rod (Greulich, 1934) and administration of hCG (Wildt &

Seager, 1978; Cline et al., 1980) or GnRH (Chakraborty et al., 1979) can induce ovulation. Ovulation stimulated with hCG or GnRH results in comparable numbers of released oocytes, regardless of whether administered in a natural or

(12)

FSH-P induced oestrus (Goodrowe & Wildt, 1987). Excessive doses can result in ovarian hyperstimulation and degeneration of oocytes. Dosages of 75-100 iu hCG give better results than 200 iu hCG as demonstrated by fewer degenerated oocytes (Goodrowe et al., 1988). More mature oocytes are obtained by administering 100 iu hCG on day 3 rather than on days 1 or 2 of natural oestrus (Donoghue et al., 1993). A combined regimen of eCG and hCG has been used to induce oestrus and ovulation in domestic cats and has resulted in the birth of kittens following laparoscopic artificial insemination (Howard et al., 1992).

Pseudopregnancy

If ovulated oocytes are not fertilised or pregnancy fails for other reasons, the female cat undergoes a short luteal phase or “pseudopregnancy”. The cat differs from other canivores in that the luteal phase of the non-pregnant cat is only approximately one half the duration of the normal gestation period (Feldman &

Nelson, 1996). Pseudopregnancy in the cat is not associated with behavioural changes or lactation as in the dog. Plasma progesterone reaches maximum levels around 20-25 days after mating (Verstegen, 1998) and begins to decline gradually from day 25 to reach basal values around 30-40 days after the first copulation (Verhage et al., 1976; Shille & Stabenfeldt, 1979; Shille & Sojka, 1995).

The anatomy of the cervix and cervical patency

Ovarian cyclicity induces considerable changes in morphology, physiology and function of the reproductive organs. The cervix, the uterus and the uterine tubes are target organs affected directly by hormonal changes during the various stages of the oestrous cycle. The cervix in the cat is oriented horizontally in straight alignment with the vestibule and vagina. The vestibule-vaginal junction (the so- called cingulum) and the anterior vagina are narrow and non-distensible, and the vagina narrows towards the cervix (Crouch, 1969; Watson & Glover, 1993; Shille

& Sojka, 1995). Examination of the vagina is only possible with the cat under general anaesthesia and with a 3-mm cystoscope equipped with a device for dilating the vaginal lumen (Shille & Sojka, 1995). Watson & Glover (1993) and Swanson & Godke (1994) report the length of the combined vestibule and the vagina to be 40 mm and 49.9±1.1 mm. The portio vaginalis uteri can be reached only with ≤1-2 mm wide catheter (Watson & Glover, 1993). The prominent dorsal median postcervical fold and the fornix located ventrolateral to the external cervical opening (Crouch, 1969) renders transcervical catheterisation difficult. The anatomical construction of the vagina and cervix in the cat tends to direct a catheter into the vaginal fornix, resulting in unsuccessful transcervical catheterisation (Hurlbut et al., 1988).

The cervix appears to be an important barrier for spermatozoa after both natural mating and artificial insemination. Intravaginal insemination with fresh semen results in conception rates of only 42.9% (3/7 cats) (Sojka et al., 1970) and 10.6%

(6/56 attempts) with frozen-thawed semen (Platz et al., 1978). Conception rates of 50% (9/18 cats) (Howard et al., 1992) and 80% (8/10 cats) (Tsutsui et al., 2001)

(13)

have been reported when laparoscopic intrauterine insemination and unilateral intrauterine horn insemination with fresh semen was used.

In the dog, results from artificially inseminated semen deposited in the vagina are inferior compared to semen deposited in the uterus (Linde-Forsberg et al., 1999: Linde-Forsberg, 2001). The number of fresh spermatozoa required for successful intrauterine insemination in the cat is 8x10⁶ (Tsutsui et al., 2001) whereas to obtain similar results 80x10⁶ spermatozoa are required for intravaginal insemination (Tanaka et al., 2000). Since surgical intrauterine insemination is not permitted in some countries, transcervical intrauterine insemination would be an alternative non-invasive method for improving the success rate of assisted feline reproduction, and is a potentially useful non-surgical technique for diagnosis and therapy of uterine diseases. However, proper devices and skilled knowledge for performing transcervical catheterisation in the cat are limited. Only two reports of transcervical catheterisation in the cat have been published and these were conducted for embryo transfer (Hurlbut et al., 1988; Swanson & Godke, 1994).

Cervical dynamics during the oestrous cycle have been studied in the bitch (Silva et al., 1995; Verstegen et al., 2001) but have not been reported for the cat.

Using vagino-uterographic contrast studies, the cervix of the bitch was seen to close 6.7±1.4 days (Silva et al., 1995) and 6.9±1.1 days (Verstegen et al., 2001) after the estimated time of the LH peak, when most bitches are still in oestrus.

Thus, cervical closure is suggested to be a limiting factor for reproductive success after natural mating or intravaginal insemination in the dog.

Sperm transport through the cervix and distribution in the female reproductive tract

The function of the cervix in restricting the entry of fluids, particles or spermatozoa into the uterus has been widely studied in several species but information for cats is not available. The cervix functions as a main barrier to spermatozoa for species with vaginal deposition of semen at natural breeding, such as primates, rabbits and ruminants. The tomcat appears to deposit semen in the vagina during natural mating (Blandau, 1973; Watson & Glover, 1993). Cat spermatozoa, therefore, have to pass through the cervix in order to reach the site of fertilization in the ampullary region of the oviduct (Van Der Stricht, 1911).

Diagnostic imaging, such as positive contrast radiography and scintigraphy are used to demonstrate the transport of radiopaque fluid and radioactive particles through the cervix into the uterus. The transport of radiopaque fluid from the vagina through the cervix has been demonstrated in the rabbit (Akester & Inkster, 1961) and the dog (Linde, 1978; Silva et al., 1995; Verstegen et al., 2001). With the use of hysterosalpingo-radionuclide scintigraphy (HERS), radiopharmaceutical particles (^99mTc labelled human albumin spheres) have been shown to migrate from the vagina into the peritoneal cavity in women (Iturralde & Venter, 1981).

With the use of scintigraphy, radiolabelled rabbit spermatozoa inseminated vaginally have been observed in the uterus within 5 min (Bockisch, 1993). During oestrus, dead spermatozoa (Noyes et al., 1958) and inert radioactive polystyrene

(14)

microspheres (Glover & Patterson, 1963) placed in the vagina have also been shown to enter the uterus in rabbits.

The localisation of spermatozoa in the female reproductive tract provides information both on cervical patency to spermatozoa and on the functions of certain regions of the reproductive tract. Structures acting as sperm barriers or as sperm reservoirs after insemination or natural mating have been defined in many species of animals. The cervix is the primary barrier for spermatozoa in animals with vaginal semen deposition at natural breeding, whereas the uterotubal junction (UTJ) serves to further restrict sperm access to the uterine tubes (Scott, 2000). The cervix, the UTJ and the lower isthmus are reported to be sites for sperm storage in the cow (Hunter et al., 1991) and the pig (Flechon & Hunter, 1981; Mburu et al., 1997) and the uterine tube, regardless of region, forms a functional sperm reservoir in vitro in the bitch (Pacey et al., 2000). However, there are no reports on sperm distribution in the cat.

Uterine morphology examined by hysterography and histology

Changes in uterine morphology are also dependent on ovarian activity during the various stages of the oestrous cycle. Laparoscopy has been used to observe changes in the uterus in the cat (Wildt & Seager, 1980) but diagnostic imaging is an alternative, non-invasive, technique. Few diagnostic imaging studies have been conducted to demonstrate the normal reproductive organs of the cat due to their diminutive size. Using radiography, with the aid of a paddle to compress the caudal abdomen, the postpartum uterus can only be observed for up to six days after parturition (Ferretti et al., 2000).

Hysterography as a method for diagnosing uterine disorders has been reported in the dog (Funkquist et al., 1985), but little data is available on the hysterographic appearance in the cat. Hysterographic appearances of cats in oestrus and interoestrus have been revealed in only one previous study (cited by Johnston et al., 2001). The physiological variations in uterine morphology as seen in the hysterograms should be confirmed by a histological study: a thorough understanding of the physiological changes is a prerequisite for accurate diagnosis of pathological changes in the uterus. The histological characteristics of the endometrium in cats have been described during oestrus (West et al., 1977); early pregnancy (Roth et al., 1995); pseudopregnancy (Boomsma et al., 1991); as well as the histopathological appearance of endometrial hyperplasia, a common uterine disorder in cats over five years and in cats given exogenous progestagens (Dow, 1962; Lawler et al., 1991; Potter et al., 1991).

Degeneration and regeneration are common features involved in epithelial development, which can be defined by the mitotic activity of the cells.

Proliferative cell nuclear antigen (PCNA) is commonly used to determine endometrial hyperplasia and adenocarcinoma in humans (Ito et al., 1993). Thus, immunohistochemical staining of the endometrial epithelial cells with PCNA might be a useful tool for defining degrees of endometrial proliferation. Mitotic figures in the endometrium during normal oestrous cycle in cats have been

(15)

described (Dawson & Kosters, 1944), but examination using PCNA has not been performed in cats.

(16)

Aims of the study

The overall aim of the present work was to increase knowledge about reproductive physiology in the female domestic cat. Special attention was directed to the morphology and function of the cervix, the uterus and the uterine tubes during different stages of the oestrous cycle, after natural mating, as well as under pathological conditions. The specific aims of the study were to:

• develop techniques for studying cervical patency and uterine appearance of the female tubular genital tract;

• examine patency of the cervix to fluids, particles and spermatozoa;

• demonstrate the distribution of spermatozoa in the female tubular genital tract after natural mating and evaluate techniques used for studying sperm transport; and

• describe differences in the hysterographic appearance during various stages of the oestrous cycle, after MPA treatment and under pathological conditions, and to relate these findings to the histological features of the endometrium.

(17)

Materials and Methods

Animals and general management

One hundred and twenty-one (121) female domestic cats were included in the studies (Table 1). Thirty of the female cats and one male cat were research colony cats (Papers II & IV). Ninety-one of the female cats were privately owned and submitted for routine spaying at the Small Animal Hospital, Chulalongkorn University, Bangkok, Thailand (Papers I & III). Six of the 30 female colony cats aged 6-9 years were housed in a group under an artificial light schedule (16L: 8D) at the Department of Obstetrics and Gynaecology, Swedish University of Agricultural Sciences (SLU), Uppsala, Sweden (Paper II). Twenty-four of the 30 female colony cats were maintained in a group with a male cat in a separate wire enclosure adjacent to the females at the Faculty of Veterinary Science, Chulalongkorn University, Bangkok, Thailand (Paper IV). The colony cats kept in Thailand were exposed to an artificial light schedule (14L: 10D), in addition to ambient light available from the windows. All colony cats were fed a commercial diet and had free access to water.

Table 1. Numbers of colony and privately owned female cats in each experiment by stage of oestrous cycle, MPA-treatment and pathological conditions

Colony

cats Privately owned cats Paper

Follicular stage

Inactive stage

Follicular stage

Luteal stage

Postpartum stage

MPA treated

Pathological conditions I

(n=17) 9 3 2 3

II

(n=6) 6 III

(n=74) 20 9 18 12 12 3

IV

(n=24) 24

Assessment of the reproductive cycle

Stage of reproductive cycle was assessed by reproductive history (Papers I-IV), behavioural observations (Papers II & IV), vaginal cytology (Papers I & II), hormonal assays (Papers II-IV), ovarian examination during ovariohysterectomy (Papers I, III & IV), and histological sections of the ovaries (Paper III). The cats were considered to be in behavioural oestrus when they exhibited oestrous behaviours, such as calling, rubbing, rolling, lordosis, treading of the hind legs and crouching to the floor. Vaginal cytology was performed using a 2-mm diameter cotton swab (Förbandsmaterial AB, Partille, Sweden) moistened with physiological saline to obtain cells from the dorsal wall of the cranial vagina. The vaginal cells were smeared onto a glass slide and stained with Hemacolor (E.

Merck, Darmstadt, Germany). A vaginal smear with a clearing of the background,

(18)

a reduction of cellular debris and a proportion of superficial cells of 80% or more was considered as an oestrous smear.

Blood was collected via cephalic venipuncture at the time of examination (Paper II) or ovariohysterectomy (Papers III & IV). The serum was separated by centrifugation and stored at -20˚C until assayed. The hormonal analysis was performed at the Department of Clinical Chemistry, Swedish University of Agricultural Sciences, Uppsala. The serum was analysed for oestradiol-17β by radioimmunoassay using enhanced luminescence (Amerlite Estradiol 60-assay, Ortho-Clinical Diagnostics, Amersham, UK) (Papers II-IV). Progesterone was determined by luminescence immunoassay (Immulite Progesterone, Diagnostic Products Corporation, LA, USA). The inter- and intra-assay coefficients of variation for oestradiol-17β were 30.5% and 17.4% at 7.1 pmol/L; 8.6% and 2.3%

at 47.0 pmol/L; and 19.7% and 8.0% at 128.3 pmol/L. The minimal assay sensitivity of oestradiol-17β was 4.2 pmol/L. The inter- and intra-assay coefficients of variation for progesterone were 13.2% and 2.5% at 2.3 nmol/L;

6.7% and 1.2% at 24.6 nmol/L; and 3.6% and 1.7% at 54.2 nmol/L. The minimal assay sensitivity of progesterone was 0.6 nmol/L.

The ovaries were examined after ovariohysterectomy for the presence or absence of follicles and CH or CL. Activity of CL was assessed by the degree of vacuolation of the luteal cells in the histological sections of the ovaries, indicating cell degeneration according to Dawson (1941) (Paper III).

Induction of oestrus and ovulation (Paper IV)

Oestrus in 24 female cats was induced by an intramuscular administration of 100 iu eCG (Folligon, Intervet International Inc., Boxmeer, The Netherlands) during anoestrus or interoestrus. Successful oestrous induction was assessed by signs of oestrous behaviour during 5-7 days after the eCG administration. Ovulation was induced by natural mating with a male cat four times within one hour. Successful intromission was assessed by the expression of a characteristic vocalisation by the female cat, subsequent disengagement of the male and typical female post-coital behaviour (Axnér & Linde-Forsberg, 1998). Ovulation was confirmed by the presence of CH and CL observed after surgery and a rise in serum progesterone concentration.

Sedation and anaesthesia

Food was withheld for at least 8 hr before sedation and anaesthesia. Sedation was performed using 100-120 µg kg^-1bw medetomidine HCl (Domitor, Orion Pharma AB Animal Health, Espoo, Finland) intramuscularly (Paper II). At the end of the examination, Atipamezol HCl (Antisedan, Orion Pharma AB Animal Health, Espoo, Finland) was administered intramuscularly in half the volume of the previously given dose of medetomidine. The cats were pre-medicated with 0.04 mg kg^-1 bw atropine sulphate (Atropine A.N.B. Laboratories, Bangkok, Thailand) (Paper I). General anaesthesia was induced with an intramuscular injection of 3 mg kg^-1bw xylazine HCl (Rompun, Bayer Korea, Seoul, Korea) and 10 mg kg^-1

(19)

bwketamine HCl (Calypsol, Gedeon Richter, Budapest, Hungary (Paper I); Keta- hameln, Astrapin Pharma GmbH & Co.Kg, Hameln, Germany (Papers III & IV)).

Investigation of the cervix

Transcervical catheterisation

A transcervical catheter consisting of an inner and an outer polypropylene tube was developed to fit the anatomy of the cat vagina and cervix (Paper I). The outer, vaginal, catheter was designed from a polypropylene urinary catheter (2.8 mm diameter) (Clay Adams, Sparks, USA), by closing one end with heat, and making a new opening 8 mm from the tip. A 3.5 French tomcat catheter (Sherwood, St.Louis, USA) 140 mm in length and 1-mm in diameter was used as an inner, cervical, catheter. The new opening in the vaginal catheter was fashioned in such a way that it directed the inner catheter towards the cervical opening.

With the anaesthetised cats lying in lateral recumbency, transcervical catheterisation was performed by inserting the outer catheter into the vagina until no further cranial movement could be achieved. The inner catheter was thereafter inserted through the outer catheter and pushed blindly through the cervix into the body of the uterus (Papers I & II). The success of the transcervical catheterisation was assessed during surgery when the tip of the catheter was palpated in the body of uterus and green food colour mixed with 1 ml of 200 000 units penicillin G (Penomycin M and H Manufactering, Bangkok, Thailand) was seen in the uterus (Paper I). Transcervical catheterisation was performed in cats at different stages of the oestrous cycle to assess the success rate of the method (Paper I) and to introduce a contrast medium into the uterus to perform hysterography (Paper III).

Assessment of cervical patency

Cervical patency to contrast fluid and a radiopharmaceutical medium

The cervical patency was examined in 17 privately owned cats in various stages of oestrus at the time of routine ovariohysterectomy. The exact day of the oestrous cycle was not known for all cats (Paper I). Cervical patency was studied in a non- ovulatory oestrous cycle in six colony cats examined at 2-day intervals at early oestrus, mid-oestrus, late oestrus and during interoestrus (Paper II). Additional examinations were performed after late oestrus until the contrast fluid remained in the vagina. Cervical patency was assessed using radiography (Papers I & II), fluoroscopy (Paper II) and scintigraphy (Paper II). Diatrozate 76% (Urografin, Schering AB, Berlin, Germany) (Paper I) or Iohexol 300 mg Iodine/ml (Omnipaque, Nycomed AB, Roskilde, Denmark) (Paper II) were used as the contrast medium. Radiopharmaceutical medium was prepared by reconstituting a vial of human serum albumin particle (HSA) with 2 ml of 400 MBq eluted technetium 99m (^99mTc). To localise the position of the uterus in relation to the kidneys and other internal organs, 0.5 ml eluted ^99mTc (50 MBq) was administered intravenously.

The contrast fluid and the radiopharmaceutical medium were infused slowly through a 3.5-mm French tomcat catheter placed in the cranial vagina with the cats

(20)

lying in lateral recumbency (Paper I), or in dorsal recumbency and with the hindquarters elevated at an angle of approximately 15-degrees (Paper II).

Radiographs ((Picker, Picker International Inc., Cleveland, USA) (Papers I & III) or Siemens-Elema, München, Germany (Paper II)) were taken in lateral projection after 0.5 ml Urografin was infused into the cranial vagina and the hindquarters of the cats were elevated for 5 min (Paper I) or in ventrodorsal projection at 1-, 3- and 5 min after Omnipaque infusion (Paper II). Additional lateral exposures were taken as required. Fluoroscopic recordings (Siemens-Elema, München, Germany) were taken at the time of Omnipaque infusion (Paper II). During mid-oestrus and interoestrus, the cats were examined with a gamma camera (Picker SX-300; Picker International Inc., Cleveland, USA) equipped with a LEGP collimator and a dedicated computer system (NUD, Nuclear Diagnostics, Stockholm, Sweden) (Paper II).

Through the tomcat catheter, 0.2 ml of the radiopharmaceutical medium (^99mTc- HSA) (40 MBq) was introduced into the cranial vagina. The cervix was defined as open when the contrast fluid or the radiopharmaceutical particles was seen to enter the uterus. The cervix was determined as closed when the contrast fluid or the radiopharmaceutical medium remained in the cranial vagina. The relationship between cervical patency to the contrast medium and oestrous behaviour, cornification of the vaginal cells and the serum oestradiol-17β concentration were evaluated (Paper II).

Cervical patency to spermatozoa (Paper IV)

The ovaries and reproductive tracts of 24 female cats were surgically removed at 30 min (n=6), 3 hr (n=6), 48 hr (n=6) and 96 hr (n=6) after natural mating. The females ovariohysterectomised at 30 min were mated only once, whereas the other females were mated four times in one hour. The reproductive tracts were divided into seven segments on each side: infundibulum, ampulla, isthmus, UTJ, cranial and caudal uterine horn, and uterine body. The vagina and the lumina of the segments from one side were flushed with 0.5 ml phosphate-buffered saline solution (PBS). The contralateral side of the reproductive segments was fixed immediately without prior flushing and these fixed reproductive segments were then further processed for routine histology. The differences in the numbers of spermatozoa in the vagina and each reproductive segment were evaluated.

Investigation of sperm distribution in the female reproductive tract (Paper IV)

Sperm recovery by flushing

Eighteen female cats were mated four times in one hour with the same male cat and were ovariohysterectomised at 3 hr (n=6), 48 hr (n=6) and 98 hr (n=6). Six females were mated only once and were submitted for ovariohysterectomy 30 min later. Each reproductive tract was divided into seven segments. The infundibulum was separated at the end of the conical shaped tubal segment. The ampulla was differentiated from the isthmus by its convoluted shape and slightly larger diameter. The UTJ comprising 0.3 cm of the tip of the uterine horn and 0.2 cm of

(21)

the caudal isthmus was excised. The remainder of the uterine horn was divided into two equal segments, defined as the cranial and the caudal uterine horn. The body of the uterus was divided longitudinally into an equal left and right part. The vagina and the lumen of each segment of one side were flushed with 0.5 mL PBS.

The flushings were collected into separate plastic Eppendorf vials.

Centrifugation at 1500 g for 5 min concentrated the spermatozoa in the flushings.

Approximately half the volume of the supernatant was discarded. Of the resuspended samples, 5µL was placed on a glass slide, covered with an 18x18 mm cover slip and examined under a phase contrast microscope at x100 magnification to calculate total numbers of spermatozoa in the flushings from each segment and from the vagina.

Sperm numbers from tissue sections

After flushing, the seven flushed segments and the corresponding contralateral non-flushed segments were immersed in 3% neutral buffered formalin solution.

Each segment was cut transversely into four equal parts, embedded in a paraffin block and sectioned to a thickness of 5-µm. Every fifth serial section was mounted and stained with haematoxylin and eosin. From each segment, 40 sections were chosen for counting of sperm numbers under a light microscope at 400x magnification.

Evaluation of techniques used for studying sperm distribution

Flushing and tissue sectioning techniques for examining sperm numbers in the female reproductive segments were evaluated by comparing sperm numbers obtained from flushed and non-flushed tissue sections.

Investigation of the uterus

Assessment of uterine morphology Hysterography

The uterine appearance of 80 cats in inactive (n=20), follicular (n=15), luteal (n=18) and postpartum (n=12) stages of the oestrous cycle, cats treated with MPA (Depo-gestin, A.N.B. Laboratory, Bangkok, Thailand) (n=12) or with uterine pathological lesions (n=3) were assessed using a positive contrast study (Papers II

& III). The cats were positioned in dorsal recumbency and the hindquarters elevated at an angle of approximately 15-degrees (Paper II) or 30-degrees (Paper III). When the cervix was shown to be functionally open, the infusion of Omnipaque into the cranial vagina was continued until the endometrial lining was observed (Paper II). Transcervical or intrauterine catheterisation was performed to introduce the contrast fluid into the uterus in cats with closed cervix (Paper III). A volume of 2-3 ml Omnipaque was infused until an efflux of the contrast was observed in the vagina (Papers II & III). Radiographic images were taken ventrodorsally and laterally immediately after the contrast infusion was completed (Paper III), and ventrodorsally at 1-, 3- and 5 min after the uterine horns were

(22)

entirely filled with the contrast medium as observed with fluoroscopy (Paper II).

The hysterographic appearances of cats in different groups were described (Papers II & III).

Nomenclatures and measurements to describe the hysterographic appearances were established based on assessment of the shapes of the uterine horns and luminal cavity and the characteristics of the intraluminal lining. The shape of the uterine horns was classified as being straight or curved and that of the luminal cavity as being straight, wavy or coiled. The intraluminal lining was classified as being smooth, or as irregular when the lining of the uterine horns showed filling defects in the contrast medium. The degree of coiling and waviness were expressed as a ratio between the number of waves or coils per 2 cm of uterine length. The amplitude of the coiling was defined as the distance between peaks of the contralateral coil (Paper III).

Ultrasonography (Paper II)

During oestrus in six of 80 female cats, ultrasonography of the uterus was performed once with ultrasound (Apogee CLA, Interspec Ins., Ambler, USA) and a 5-10 MHz changeable convex array probe (Apogee CLA, Interspec Ins., Ambler, USA). The endometrial appearances were described. The diameter of the uterine horns was measured, as were the cystic formations in the endometrial wall.

Histology and immunohistochemistry (Paper III)

After hysterography 74 of 80 cats were ovariohysterectomised and the uteri measured for length and diameter. Subsequently, the uteri were fixed in 10%

buffered formalin solution, embedded in paraffin, sectioned and stained with haematoxylin and eosin and the endometrium was evaluated histologically for lesions in the luminal epithelial, subepithelial and glandular layers. An image analyser (Microphot-FXA, Nikon Inc., Tokyo, Japan) equipped with a computer system (Easy Image Measurement, Bergström Instrument AB, Solna, Sweden) was used for measuring endometrium thickness, myometrium thickness, luminal epithelial cell height and glandular epithelial cell height.

The endometrial cell activity in cats at various stages of the oestrous cycle, cats treated with MPA and cats with uterine pathological conditions was investigated by immunohistochemical labelling with avidin-biotin peroxidase complex (ABC), according to the procedures described previously (Simoes et al., 1994). Non- specific binding sites were decreased by incubating the tissue with 10% bovine serum albumin (Fluka, Buchs, Switzerland): deparaffinised and rehydrated sections of the uteri were incubated with monoclonal anti-mouse proliferating cell nuclear antigen (PCNA) (Dako, Glostrup, Denmark) (1:200). Thereafter, sections were treated with biotinylated rabbit anti-mouse IgG antibody (Dako, Glostrup, Denmark) (1:400). The immunostaining was developed with 0.05% 3,3- diaminobenzidine tetrahydrochloride in 0.01 M Tris-HCl, pH 7.6 (DAB) (Sigma- Aldrich, St. Louis, USA) and counterstained with Mayer’s haematoxylin. The proliferation index (PCNA index) of the luminal and glandular epithelium was calculated from the percentage of nuclear positive immunoreactive cells from 500 cells in five randomly selected fields.

(23)

Assessment of uterine motility Fluoroscopy and scintigraphy (Paper II)

Movement of the uterus was observed with fluoroscopy and scintigraphy in a non- ovulatory oestrous cycle in six cats. Fluoroscopy was performed at 2-day intervals at early oestrus, mid-oestrus and late oestrus and fluoroscopic recordings were taken continuously from completion of the Omnipaque infusion into the cranial vagina to 5 min after the contrast medium filled the uterine horns. Scintigraphy was performed once during mid-oestrus. Simultaneous with the ^99mTc-HSA infusion, dynamic hysteroscintigraphy was obtained with a gamma camera at 1- frame/sec in 128 x 128 x 16 matrix for 180 sec. The other 30 sec dynamic images (1 frame/sec) were taken at 5 and 10 min after ^99mTc-HSA infusion. The intrauterine transcornual movement of the contrast fluid and the radiopharmaceutical medium was described according to the contraction pattern of the uterus.

Statistical analyses

Data generated in Papers III & IV were analysed with the Statistical Analysis Systems software (Vers. 8, SAS Institute Inc, Cary, NC, USA). Analysis of variance (ANOVA) was applied to quantitative data using general linear model (GLM). Normal distribution of residuals from ANOVA was tested using the UNIVARIATE procedure. Variables that were not normally distributed were transformed into 10-logarithms or inverse numbers (Paper III). The Tukey- Kramer test was used to compare mean values and the FREQ procedure was used to analyse the frequency distribution of hysterographic features and histological characteristics in each group of cats (Paper III). Differences in mean scores between groups for categorical data in which the normality could not be obtained after transformation were analysed by the NPAR1WAY procedure (Wilcoxon’s rank-sum test) (Papers III & IV). A p-value ≤ 0.05 was considered as statistically significant.

(24)

Results

Clinical observations and hormonal analyses

The duration of natural oestrous behaviour of six cats ranged between 5 and 8 days, and the duration of the maximum vaginal cornification ranged between 4 and 8 days (Paper II). Onset of oestrous signs of 24 induced cats was observed on Days 3-7 after eCG administration (Paper IV). The eCG treatment induced an average of 10.4±6.3 mature follicles (≥2 mm) per cat (n=24) (Paper IV). The mean serum concentration of oestradiol-17β during interoestrus was 13.2 pmol/L (range 8-26 pmol/L) (n=22) (Papers II & III). Natural behavioural oestrus was first detected when serum concentration of oestradiol-17β was greater than 80 pmol/L (n=6) (Paper II): peak serum concentration of oestradiol-17β occurred within 4 days after the onset of oestrous behaviour (Paper II). A mean oestradiol-17β concentration of 47.7 pmol/L (range 14-113 pmol/L) was observed in cats in which mature follicles were present in the ovaries (≥ 2 mm in diameter) without knowing the accurate day of oestrus (n=2) (Paper III). The means and ranges of oestradiol-17β concentrations analysed from cats in a natural oestrous cycle (Paper II) and an induced oestrous cycle (Paper IV) are illustrated in Table 2. Oestradiol- 17β concentrations approached basal levels on Day 8 of natural oestrus whereas oestrous behaviour subsided from Day 5 (2/6 cats), Day 6 (1/6 cats), Day 7 (2/6 cats) or Day 8 (1/6 cats) (Paper II).

Table 2. Oestradiol-17β concentrations (pmol/L) in serum of cats in natural and in induced oestrus. Means (ranges)

On Day 3 of the eCG-induced oestrus, four intromissions within one hour successfully induced ovulation in 12 females (Paper IV). In 12 cats with an eCG induced oestrus, 123 of 140 mature follicles were ovulated as confirmed by the presence of CH or CL (87.9%) (Paper IV). Progesterone was at basal level on Day 3 of eCG-induced oestrus (n=12) (Paper IV), the level was still basal at 48 hr after mating (Day 5 of oestrus), except in one cat with a high serum progesterone level (46.5 nmol/L). At 96 hr after mating (Day 7 of eCG-induced oestrus), serum progesterone was 74.4 nmol/L (range 31.2-127.2 nmol/L) (n=6) (Paper IV).

Serum concentration of oestradiol-17 β (pmol/L) Day of oestrous

cycle Numbers

of cats Natural oestrus Induced oestrus

2 6 125.5 (85-180) -

3 12 - 119.2 (53-180)

4 6 91.2 (30-180) -

5 6 - 84.7 (20-180)

6 6 36.8 (20-62) -

7 6 - 25.7 (8-42)

8 4 21.8 (17-30) -

10 2 23.0 (18-28) -

(25)

Techniques used for the studies

Transcervical catheterisation and cervical patency

The distance from the vulva to the cranial vaginal fornix of the cats was assessed as being 45-60 mm when measured with the aid of the 3.5 French tomcat catheter (Paper II). Urografin (0.5 ml) deposited vaginally through a tomcat catheter was not observed entering the uterus, even in oestrous cats when they were lying in lateral recumbency (Paper I). When the cats were positioned in dorsal recumbency with the hindquarters elevated at an angle of 15-degrees, passage of the Omnipaque and ^99mTc-HSA was observed with fluoroscopy and a gamma camera during oestrus (Paper II).

Slow infusion of the contrast medium was necessary to prevent backflow. When the cervix was functionally open, a continued infusion of 2-4 ml of Omnipaque, including an amount of 0.5-1.0 ml backflow, would fill the uterine horns (Paper II). A similar volume of Omnipaque (2-3 ml) was sufficient to fill the uterine horns using transcervical catheterisation, with or without cervical manipulation through laparotomy or intrauterine deposition using an intravenous catheter at the time of surgery (Paper III).

Sperm recovery techniques (Paper IV)

There were no differences in sperm numbers in the tissue sections between the flushed and the non-flushed segments of the reproductive tract in the cats ovariohysterectomised at 30 min, 3 hr, 48 hr and 96 hr after mating (p>0.05).

During early sperm transport (at 30 min and 3 hr after mating), more spermatozoa were recovered by flushing than observed in the tissue sections, whereas, after ovulation (at 48 hr and 96 hr after mating) more spermatozoa were observed in the tissue sections than were recovered by flushing.

Success of transcervical catheterisation

Transcervical catheterisation with the specially designed catheter was successful in 20/29 cats in the inactive stage of the oestrous cycle; 9/12 in the follicular stage:

9/20 in the luteal stage; 15/15 in the postpartum stage; 12/12 cats treated with MPA; and 2/3 cats with pathological lesions in the uterus (Papers I & III). In 32 cats, blind transcervical catheterisation was successful in 9/29, 5/12, 4/20, 7/15, 5/12 and 2/3 cats, in the respective groups (Papers I & III). Blind insertion of the inner catheter without manipulation of the cervix through laparotomy was successful in 35.2% of cases (32/91 attempts) (Papers I & III).

Cervical patency during oestrus

Cervical patency to contrast fluid and a radiopharmaceutical medium (Paper II)

During interoestrus, Omnipaque or ^99mTc-HSA remained in the vagina. In some of the oestrous stages when the cervix was open, vaginally deposited contrast fluid or

(26)

radiopharmaceutical medium was rapidly transported through the cervix into the uterus. The cervical patency dynamics varied among the cats. In 3/6 cats, transcervical transport of the Omnipaque was identified in all stages of oestrus; in 1/6 cats during mid-oestrus, late oestrus and 1 day after oestrus; and in 2/6 cats only during late oestrus.

The patency of the cervix in all six cats occurred after the peak and during declining levels of serum oestradiol-17β, the concentration ranging from ≥180 to 17 pmol/L. The cervix was open at the time of maximal cornification of the vaginal cells in 13/16 observations (>80%). The closure of the cervix, characterised by the Omnipaque remaining in the vagina, occurred following the end of the behavioural oestrus and the phase of maximum cornification of the vaginal epithelium in 5/6 cats. In the remaining cat, the cervix was still open after this stage, on Day 8 of oestrus. Rapid transportation of particles from the vagina through the cervix into the uterus in 3/5 cats during mid-oestrus was determined by scintigraphic examination.

Cervical patency to spermatozoa (Paper IV)

The cervix was found to act as a sperm barrier in the cat. At 30 min after natural mating, only one quarter (1.1x10⁴/4.5x10⁴) of the mean total number of spermatozoa recovered from the reproductive tract by flushing was found in the uterus and the uterine tubes. Differences in sperm numbers (p<0.05) were found between the vagina and the uterine segments including the UTJ flushed at 30 min and 3 hr following mating (n=12). The total number of spermatozoa recovered by flushing throughout the reproductive tract at 30 min after one mating did not differ from that at 3 hr after the fourth mating (p<0.05). Spermatozoa were still detected in the vagina in 5/6 cats at 48 hr after mating and 1/6 cats at 96 hr after mating.

Localisation of spermatozoa in the female reproductive tract (Paper IV)

There was a considerable variation among cats both within and between groups in the numbers of spermatozoa recovered by the flushing procedure and from the tissue sections. The numbers of recovered spermatozoa decreased with time after mating (p<0.05). The highest proportion of recovered spermatozoa at 30 min, 3 hr, 48 hr and 96 hr after mating was detected in the vagina. Before ovulation (determined at 30 min and 3 hr after mating), the majority of spermatozoa were found in the vagina and the uterine segments: after ovulation (determined at 48 hr and 96 hr after mating), higher numbers of spermatozoa were present in the uterine tubal segments than before ovulation. During the time before ovulation, sperm numbers recovered by flushing from the vagina, the uterine segments (including the UTJ) and the uterine tubal segments differed (p<0.05).

In the histological sections, the majority of spermatozoa were located in groups in the epithelial crypts, with only a few residing in the lumina of the reproductive tract. There were no differences in numbers of recovered spermatozoa in the tissue sections between the flushed and the non-flushed segments of the reproductive

(27)

tract in any of the four groups of cats (p>0.05). At 30 min and 3 hr following mating, the majority of the spermatozoa were found in the uterine segments (including the UTJ) (p<0.05), whereas at 48 hr and 96 hr they were mostly located in the uterine tubal segments (p<0.05). Higher sperm numbers were detected in the ampulla and isthmus after ovulation than before ovulation (p<0.05). However, in the isthmus, the sperm numbers recovered by both flushing and counting in tissue sections, were highest at 48 hr after mating (p<0.05).

Uterine morphology assessed by hysterography and ultrasonography

With positive contrast techniques, the uterine appearance in cats at various stages of the oestrous cycle, in cats treated with MPA, and in cats with pathological lesions in the uterus, differed (Papers II & III). Thus, criteria for hysterographic descriptions were defined. Straight shape of the uterine horns with a straight luminal cavity and a smooth inner luminal contour was indicative of cats in the inactive stage of the oestrous cycle, whereas a wavy luminal shape of the uterine horns was characteristic for cats in the follicular stage of the oestrous cycle. In the luteal stage and during the various phases of luteal activity, the luminal cavity of the uteri varied in shape between being straight, irregular wavy and coiling. A coil-shaped uterine lumen seen in the MPA-treated and pathological groups was suggested as characteristic for a progestagenic effect as observed in some cats in the luteal group. In the colony of old cats kept in Uppsala, Sweden (Paper II), the hysterograms examined during oestrus were described as a spiral- (2/6) and a tight spiral- (2/6) uterine horn based on degrees of waviness of the uterine cavity whereas coiled luminal shape of uterine horns with irregular filling defects was defined as a corkscrew appearance (Paper II). Ultrasonography demonstrated a few 1 mm cystic changes in the endometrial wall of one of the cats showing spiral- shaped uterine horns with a smooth intraluminal lining and in both of the cats that had a tight spiral-shaped uterus with a smooth inner contour. Multiple 1-2 mm cystic endometrial changes were observed in both of the cats with a corkscrew appearance of the uterus and distinct irregular filling defects. Waviness and coiling of the uterine lumen was related to a proliferation of the endometrial glands, whereas irregular filling defects in the contrast medium were indicative of endometrial cystic changes as confirmed by the histological study.

Uterine morphology assessed by histology and immunohistochemistry (Paper III)

The histological characteristics of the endometrium in cats during various stages of oestrus and in cats with pathological conditions differed. The endometrium of cats in the inactive and postpartum stages of the oestrous cycle was characterised by a single luminal and glandular epithelial cell lining. Pseudostratification of the luminal epithelial cells was characteristic for the endometrium during the follicular stage of the oestrous cycle. During the luteal stage, both the luminal and glandular epithelium were pseudostratified. Hyperplastic changes of the luminal and

(28)

glandular epithelium were the predominant characteristics of histopathological features seen in the MPA-treated and pathological groups.

Length of the uterine horns measured from gross specimens was found greatest in the luteal stage of the oestrous cycle although the differences observed between the stages of the oestrous cycle and between normal and pathological uterine horns were not significant. The outer diameters of the uterine horns measured from gross specimens and from hysterograms between groups of cats were not significantly different (p>0.05). The luminal diameters measured from hysterograms decreased in correspondence to the increase of endometrial and myometrial thickness measured microscopically. The endometrium thickness was greatest in the pathological group whereas the myometrium was thickest in the MPA treated group. The luminal epithelial cells were highest in the follicular stage whereas the glandular epithelial cells were highest in the luteal group.

The PCNA immunostaining cells showed a distinct brown nuclear positive colouration in both luminal and glandular epithelium, indicating mitotic activity in these cells. Between the groups of cats, the percentage of the positive cells in the luminal and glandular epithelium did not differ (p>0.05). A large variation in the PCNA index was noted within the groups of cats.

Uterine motility assessed by fluoroscopy and scintigraphy (Paper II)

Under the fluoroscope, uterine contractions were observed in both ascending and descending directions, flushing the Omnipaque back and forth between the uterine horns and causing reflux of contrast fluid out of the uterine horns through the cervix into the vagina. In most observations, the contrast fluid remained in the uterus for 5 min after infusion. Fluoroscopy revealed frequent and strong uterine contractions. The dynamic images of the hysteroscintigraphy displayed a migration of ^99mTc-HSA into both uterine horns immediately after deposition in the cranial vagina. Movement of the radiopharmaceutical medium within and between the uterine horns was observed. The radioactivity remained during the entire 10 min observation period. In one cat after 3 min, more ^99mTc-HSA was retained in the left uterine horn than was retained in the right uterine horn.

(29)

General discussion

Clinical findings

Female cats enter the follicular phase abruptly. A rapid rise of oestradiol-17β secreted by growing follicles at the onset of the follicular phase induces distinct changes in reproductive behaviour, vaginal cornification and opening of the cervix. Ovarian activity is recognised initially by an overt sexual behaviour: the female attracts the male by calling, rolling and treading of the hind legs. In some cases (3/6 cats) in this study, behavioural changes were found to precede changes in vaginal cornification (Paper II), similar to previously reported by Shille et al.

(1979). Thus, vaginal cytology is not a reliable tool for predicting the onset of oestrus in the cat but is advantageous for determining oestrus in timid cats that fail to exhibit obvious sexual behaviour (Shille et al., 1979; Shille & Sojka, 1995).

At the onset of the follicular phase, a rapid rise of oestradiol-17β was observed from a base line of 13.2 pmol/L to a surge of 125.5 pmol/L on Day 2 of natural oestrus (Paper II), and of 119.2 pmol/L on Day 3 of induced oestrus (Paper IV).

Thereafter, oestradiol-17β gradually decreased to approach baseline on approximately Day 8 of the natural oestrous cycle or on Day 7 of the induced oestrous cycle while the oestrous behaviour of most of the cats disappeared. The results were in accordance with those of Wildt et al. (1981), in that the mean concentration of oestradiol-17β is above base line on the first day of oestrus, remains elevated throughout oestrus and returns to baseline during the last 24 to 48 hrs of oestrus, or after the onset of ovulation.

Ovulation was successfully induced after four natural matings within one hour as confirmed by the presence of CH or CL in the ovaries (n=12) (Paper IV). The regimen of an eCG-induced oestrus followed by natural mating used in this study resulted in a high proportion of ovulations (87.9%) similar to the proportion of oocytes recovered after eCG-induced oestrus and hCG-induced ovulation (91.4%) as reported by Goodrowe et al (1988).

Transcervical catheterisation

The specially designed transcervical catheter developed in this study enabled the introduction of contrast medium into the uterus of the cats during all stages of the oestrous cycle (Papers I & III). Thus, positive contrast study of the uterine appearance in the cat is also possible when the cervix is closed. However, there was a risk of damage to the vagina and cervix when the catheter was introduced blindly into the vagina, which is narrow, and non-distensible (Swanson & Godke, 1994). Measurements with the tomcat catheter determined the length of the combined vestibule and vagina to be 45-60 mm, similar to the post-mortem findings of Watson and Glover (1993). A single-sized vaginal catheter might not fit properly into the ventral vaginal fornix of all cats since the anatomy of the posterior reproductive tract varies between individuals. It was found that the vaginal catheter could not always be properly placed in the cranial vagina because

(30)

the portio vaginalis uteri created a too shallow ventral vaginal fornix;

consequently, the tomcat catheter did not attain the proper angle towards the cervical canal (K.Chatdarong, E. Axnér & C. Linde-Forsberg, unpublished). The overall success rate of introducing the inner catheter through the cervix (35.2%) (Papers I & III) was lower than the success rate reported in Paper I (76%), presumably as a result of more variation in anatomic structure when more cats were used. It was apparent that to perform transcervical catheterisation good knowledge of the anatomy of the reproductive tract and the individual variation between females as well as the skill of the veterinarian are required.

Cervical patency

Cervical patency could be demonstrated using vaginal deposition of a contrast medium, as well as a radiopharmaceutical, through a 3.5 French tomcat catheter in cats lying in dorsal recumbency and with the hindquarters elevated (Paper II) but not when in lateral recumbency (Paper I). It might be that cats lying in dorsal recumbency provided better pooling of the contrast medium around the opening of the cervical canal, compared to when the cats were positioned in lateral recumbency (Paper I). Similar observations have been reported in the bitch (Linde, 1978).

In most cases (13/16 observations), the cervix was found to be patent in relation to the presence of maximal cornification of the vaginal cells and following the peak of the serum concentrations of oestradiol-17β (Paper II). In bitches, the cervix is functionally open during declining serum concentrations of oestradiol- 17β and prior to cytological metoestrus (Silva et al., 1995). The cervical dynamics appeared to be a delayed effect of the high serum oestradiol-17β in the cat as well as in the bitch. The time when the cervix was open varied between cats in this study and was seen either only during late-oestrus, during midoestrus and late- oestrus, or throughout entire oestrus (Paper II). However, it was shown that the cervix in some of the cats did not permit entry of contrast medium even when full cornification of the vaginal cells was evident (Papers I & II).

Transport of fluids, particles and spermatozoa through the cervix

The mechanism by which fluids, particles and spermatozoa are transported through the cervix is not clearly understood but is thought to involve muscular activity of the vagina, cervix and uterus (Harper, 1988). Rapid transport of the contrast fluid, the inert radiopharmaceutical particles and the spermatozoa from the vagina into the uterus of the cats, was observed by fluoroscopy, scintigraphy and flushing of the reproductive tract after natural mating (Papers II & IV). These findings demonstrated that the cervix of the cats allowed fluids, particulate matter and living cells to enter the uterus during a certain period of oestrus. In the rabbit, dead spermatozoa (Noyes et al., 1958) as well as inert radioactive polystyrene microspheres (0.8 to 30 µm in diameter) (Glover & Patterson, 1963) placed in the vagina have also been shown to enter the uterus during oestrus. Fluoroscopy has

Reproductive Physiology of the Female Cat